A framework to analyze cerebral mean diffusivity using surface guided diffusion mapping in diffusion tensor imaging.

Area of Science:

• Neuroimaging
• Medical Physics
• Computational Neuroscience

Background:

• Mean diffusivity (MD) quantifies microstructural properties in Diffusion Tensor Imaging (DTI) of cortical gray matter (GM).
• Accurate mapping of MD values from volumetric DTI data to cortical surfaces is crucial for surface-based analysis.
• Existing methods like nearest-neighbor (NN) and linear sampling are susceptible to geometric distortions in diffusion-weighted images.

Purpose of the Study:

• To introduce and validate a novel Surface Guided Diffusion Mapping (SGDM) method.
• To compensate for geometric distortions affecting MD value sampling on the cortical surface.
• To compare the accuracy of SGDM against traditional NN and linear sampling methods.

Main Methods:

• Development of the Surface Guided Diffusion Mapping (SGDM) algorithm.
• Comparison of MD values sampled using SGDM, NN, and linear methods.
• Projection of tissue classification (including CSF probability) to the cortical midsurface.

Main Results:

• SGDM yielded lower Cerebrospinal Fluid (CSF) probability values compared to NN and linear methods.
• NN and linear methods overestimated MD values in regions with geometric distortion.
• Overestimated MD values by NN and linear methods were attributed to sampling CSF instead of GM.

Conclusions:

• The SGDM method effectively corrects mapping errors caused by geometric distortions in DTI.
• SGDM provides more accurate MD values for cortical surface analysis than conventional methods.
• Accurate MD mapping is essential for reliable surface-based DTI studies of brain microstructure.